Hi there,
I have my DM640's for a long long time by now. A few years back I replaced the ferrofluid ad put them on spikes. Other than that, they are as they came from the factory.
They are almost 30 years old by now and I thought it to be a good time to refurbish the crossovers.
The new parts will be a fair bit larger than what's in there now, so the idea is to build new crossovers from scratch.
I'm no audio engineer, so the idea was to keep the B&W design and just build with better parts.
This is the schematic from B&W (I corrected a typo in the schematic in the technical manual, where C3 was called C2):
Default values of components:
C1 85 uF
C2 85 uF
C3 16 uF
C4 7 uF
C5 16 uF
R1 8R2
R2 8R2
R3 1R0
L1 5 mH
L2 0.3 mH
L3 2.5 mH
L4 0.15 mH
New parts I'd though to be a good replacement:
Internal wiring will be replaced with 4mm2 copper wiring for the woofers and 2.5mm2 copper wiring for the mid-range driver and tweeter. I want to twist the wires and solder them directly to the drivers instead of using spades like stock. The only driver that won't be directly soldered is the tweeter, since they have sockets that would be hard to solder and I don't want to hack them up.
C1
To get to the 85 uF without breaking the bank I wanted to use the following caps. This would mean a 68 uF bipolar electrolytic with the remaining 17 uF from polycaps. From what I've read, the audible effects of electrolytic are less of a problem in the low end of the audio spectrum and this cap being in parallel would also make it less crytical.
My selection of caps that will be wired in parallel:
Mundorf ECap plain 68 uF
Jantzen Cross-cap 16 uF
Jantzen Cross-cap 1 UF
C2
This cap is in series in the crossover for the mid-range driver. I wanted to use poly caps for this.
My selection of caps, will be wired in parallel:
Jantzen Cross-cap 82 uF
Jantzen Cross-cap 3 uF
C3
This cap would be less critical, since it is in parallel. Since the cross-caps don't cost the bank, I chose to also use a poly cap for this one.
Jantzen Cross-cap 16 uF
C4
Cap in series for the tweeter, all poly caps to get to 7 uF with a small 0.1 uF bypass cap from the superior series.
Jantzen Cross cap 3 uF
Jantzen Cross cap 3 uF
Jantzen Cross cap 1 uF
Jantzen Superior-Z 0.1 uF as bypass cap
C5
Same reasoning as behind the C4
Jantzen Cross-cap 16 uF
Jantzen Superior-Z 0.1 uF as bypass cap
R1 and R2
Both Jantzen Superes 8,2 ohm, 10W
R3
Jantzen Superes 1 ohm, 10W
For the inductors I wanted to use air core inductors all arround. I know copper foil would be even better, but I think it would be overkill for these oldies, air core will be a nice step up.
L1 Jantzen 5 mH air core 13 AWG
L2 Jantzen 0.3 mH air core 18 AWG
L3 Jantzen 2.5 mH air core 18 AWG
L4 Jantzen 0.15 mH air core 18 AWG
What I'm conscerned about is the difference in resistance between the old inductors and the new. I have no idea how this will effect the crossoverpoints and / or slopes and how to correct the values of the other components to compensate for this or if this is even needed at all.
Hope someone can advise.
I have my DM640's for a long long time by now. A few years back I replaced the ferrofluid ad put them on spikes. Other than that, they are as they came from the factory.
They are almost 30 years old by now and I thought it to be a good time to refurbish the crossovers.
The new parts will be a fair bit larger than what's in there now, so the idea is to build new crossovers from scratch.
I'm no audio engineer, so the idea was to keep the B&W design and just build with better parts.
This is the schematic from B&W (I corrected a typo in the schematic in the technical manual, where C3 was called C2):
Default values of components:
C1 85 uF
C2 85 uF
C3 16 uF
C4 7 uF
C5 16 uF
R1 8R2
R2 8R2
R3 1R0
L1 5 mH
L2 0.3 mH
L3 2.5 mH
L4 0.15 mH
New parts I'd though to be a good replacement:
Internal wiring will be replaced with 4mm2 copper wiring for the woofers and 2.5mm2 copper wiring for the mid-range driver and tweeter. I want to twist the wires and solder them directly to the drivers instead of using spades like stock. The only driver that won't be directly soldered is the tweeter, since they have sockets that would be hard to solder and I don't want to hack them up.
C1
To get to the 85 uF without breaking the bank I wanted to use the following caps. This would mean a 68 uF bipolar electrolytic with the remaining 17 uF from polycaps. From what I've read, the audible effects of electrolytic are less of a problem in the low end of the audio spectrum and this cap being in parallel would also make it less crytical.
My selection of caps that will be wired in parallel:
Mundorf ECap plain 68 uF
Jantzen Cross-cap 16 uF
Jantzen Cross-cap 1 UF
C2
This cap is in series in the crossover for the mid-range driver. I wanted to use poly caps for this.
My selection of caps, will be wired in parallel:
Jantzen Cross-cap 82 uF
Jantzen Cross-cap 3 uF
C3
This cap would be less critical, since it is in parallel. Since the cross-caps don't cost the bank, I chose to also use a poly cap for this one.
Jantzen Cross-cap 16 uF
C4
Cap in series for the tweeter, all poly caps to get to 7 uF with a small 0.1 uF bypass cap from the superior series.
Jantzen Cross cap 3 uF
Jantzen Cross cap 3 uF
Jantzen Cross cap 1 uF
Jantzen Superior-Z 0.1 uF as bypass cap
C5
Same reasoning as behind the C4
Jantzen Cross-cap 16 uF
Jantzen Superior-Z 0.1 uF as bypass cap
R1 and R2
Both Jantzen Superes 8,2 ohm, 10W
R3
Jantzen Superes 1 ohm, 10W
For the inductors I wanted to use air core inductors all arround. I know copper foil would be even better, but I think it would be overkill for these oldies, air core will be a nice step up.
L1 Jantzen 5 mH air core 13 AWG
L2 Jantzen 0.3 mH air core 18 AWG
L3 Jantzen 2.5 mH air core 18 AWG
L4 Jantzen 0.15 mH air core 18 AWG
What I'm conscerned about is the difference in resistance between the old inductors and the new. I have no idea how this will effect the crossoverpoints and / or slopes and how to correct the values of the other components to compensate for this or if this is even needed at all.
Hope someone can advise.
Last edited:
Regarding inductors & their resistance Jantzen do masses of inductors which all look the same with the same wire gauge & mH value.....except that the way it was wound with a small air gap at the centre or a large one, or flat or stacked alters the DCR.
When I modified my speakers & cross-overs I decided to choose inductors with a DCR of 1/20 of that drivers stated impedance that the inductor was being used for.
When I modified my speakers & cross-overs I decided to choose inductors with a DCR of 1/20 of that drivers stated impedance that the inductor was being used for.
The replacement inductors should have the same DCR as the originals to avoid upsetting the 'voicing' of the speaker.
If not, the system could sound either better or worse to your ears.
To achieve the same low DCR as a laminated iron core inductor, an air core inductor needs to have many more turns of a much heavier gauge of wire, and this adds greatly to bulk and expense. I would expect that you already know this, if not you must consider whether it is worth replacing all the coils.
Oh, and if you can post photos of the existing crossover components, that would be nice.
If not, the system could sound either better or worse to your ears.
To achieve the same low DCR as a laminated iron core inductor, an air core inductor needs to have many more turns of a much heavier gauge of wire, and this adds greatly to bulk and expense. I would expect that you already know this, if not you must consider whether it is worth replacing all the coils.
Oh, and if you can post photos of the existing crossover components, that would be nice.
Hi, I put the bass crossover into a simulator with a Scanspeak woofer, and changed the DCR of the 5mh coil 3 times so you can see what difference it makes.
Top blue curve DCR = 1.2ohm, middle red curve = 1.8ohhm, bottom black curve = 2.2ohm. Basically, it changes the response, impedance and phase.
In this example, SPL variation is up to about 1.5db
You might consider keeping the originals, saving the money and avoiding the guesswork.
Top blue curve DCR = 1.2ohm, middle red curve = 1.8ohhm, bottom black curve = 2.2ohm. Basically, it changes the response, impedance and phase.
In this example, SPL variation is up to about 1.5db
You might consider keeping the originals, saving the money and avoiding the guesswork.
For that matter, the whole audio band is low in the spectrum.
No. My shower and W.C. water supplies are in parallel and I don't want anyone flushing while I'm in the shower.
The only necessary changes are to replace the existing Electrolytics with modern hi quality electrolytics. This will restore the capacitance and ESR of the caps back to near original. Anything else is a waste of $$$ - the other parts will still be OK unless abused by excessive power.
What's also worth mentioning here is that while thicker wire will be able to match the DCR, the AC losses of an air cored inductor will be greater, due to the increased proximity effect of all those extra turns layers. To get a close match at AC multi-stranded wire (kinda litz, but with thicker wire) would be required. (This is assuming the original laminated core itself isn't contributing significant losses which might be a big ask.)
How loud do you play these speakers? The stock inductors on the B&W DM series are iron cored. They will saturate under the 75w input range and distort pretty abruptly. I redid my DM580s with air core and added series R to make up for the difference. The difference was significant and the sound really opened up alot. I typically use Solen PP in series circuits and budget PP for parallel caps. I really like most DM series speakers uo to the early 90s. Yours will benefit greatly from a thorough crossover overhaul.
Also, be careful removing the crossovers from your speakers. They are going to be a press fit and glued into place, so you need to push them out from the inside to avoid breaking the terminal plate flanges.
From time to time I do like to play them loud. With 2x 390w RMS of amplification, it's hard to resist the volume knob at times 😉
I have not measured the resistance of the coils that are in there now, but to replace the 5mH inductor an air core would need to be quite large to be comparable to the ferite core that's in there now (resistance wise). Maybe a better quality iron core inductor would already do a lot better than the one that's in there now. Jantzen has a iron core inductor with 15 gauge wiring that has a DCR of 0,24 ohm, my hope is this would be an improvement over whats in there now without increasing resistance too much. The 0.3 mH will definatly be upgraded to an air core, the 0.15 mH one for the tweeter circuit already is an air core. The 2.5 I'm not sure yet what to go for, air core with bigger gauge or upgrade it to a better quality iron core.
Capacitors;
I was thinking to go with Jantzen eCaps (NPE, 'smooth' variant) for c1, and c3 and use Jantzen cross-caps for the rest of the crossover.
Since I'm replacing everything else and resistors don't cost much, I'd like to upgrade them as well. This way I can leave the stock crossovers as they are and just swap them out for the new crossover when it's finished.
By the way; I get what you say about the older DM series. I have 603 S2's in my home theater setup and while they do sound nice there, they are no competition for the 640's. A few years back I had the chance to switch to 604's for my music setup, but this didn't feel like an upgrade. They did look nice and more modern, but the sound quality is nowhere near the 640's(!)
If you go with a cored inductor, the only ones I recommend are the steel laminate type, making sure the core is sized conservatively. These tend to have predominantly 2nd order HD and saturate pretty late in the power curve. DcR will be a problem with air cores above 2mH if the wire gauge is thinner. Either way, avoid the powdered iron and ferrite inductors. They don't sound good in a speaker crossover IMO.
That's a good idea to keep the stock crossovers intact. That way you can experiment away to your heart's content on the separate, new crossovers.
By comparing one speaker fitted with the stock crossover with the other speaker fitted with your new crossover, you can readily judge the degree of improvement in sound that you have achieved.
I was looking at this one for L1 (5 mH)
https://www.soundimports.eu/en/jantzen-audio-000-5164.html
This one down for L3 (trim down to 2.5 mH)
https://www.soundimports.eu/en/jantzen-audio-000-5118.html
L2 (0,3 mH)
https://www.soundimports.eu/en/jantzen-audio-000-1313.html
And L4 (0,15 mH)
https://www.soundimports.eu/en/jantzen-audio-000-1051.html
Still have to measure the current ones with an LCR meter though to see how much different these ones are to the original inductors.
It's worth stating that resistance isn't bad. Of course, it's differences from the original that might give unexpected results, or maybe they wont.. or maybe they should be changed for a reason. It depends. Maybe using air cores would make an improvement in some cases, again it's hard to guess.
I’m also playing with the idea to 3d print a new terminal plate for 6 banana jacks, to keep the crossovers external. That way I can switch them out quite easily. I have enough space behind the speakers to put them in their own enclosure.
Have you considered redesigning the crossover with a new circuit or with new values? This is where the differences would come in.
If the existing components are working properly then you should expect to hear no difference if you replace them, if you properly match the values. In particular I don't see anything wrong with the resistors, not sure why you'd want to replace them. Same goes for the air cored inductor. If you measured the capacitors it would tell you about their condition.
It's not so much that I want to replace them persé. I want as little downtime as possible. By using new parts I can build the crossover and swap them when finished. 0 downtime. Also I can switch between the new and old crossover to A/B them.